Thomas Bürger

The paper introduces Pirat, a method for imputing missing values in proteomics data by modeling instrument limits and peptide correlations, showing improved performance over existing methods.

This study identifies phosphorylation changes in host proteins induced by HBV infection in primary human hepatocytes, revealing an intrinsic anti-viral response involving DNA repair and RNA-binding proteins.

The paper discusses practical considerations for controlling false discoveries after one-way ANOVA testing in proteomics, focusing on the interaction between FDR control and post-hoc tests.

The paper evaluates transcriptome-informed proteogenomic databases, finding increased peptide sensitivity is a statistical artifact, but transcriptome data reduces protein identification ambiguity.

This paper introduces knockoff filters for FDR control in proteomic differential analysis, comparing them to traditional methods on real and simulated data.

The paper challenges the use of target-decoy competition for FDR control in proteomics and proposes a more stable alternative based on the Benjamini-Hochberg framework.

Estimating what fraction of missing proteomics values are missing completely at random lets MCAR- and MNAR-specific imputers be combined; on simulations this beats standard imputation.

The paper introduces a new statistical test for differential protein abundance in proteomics data, showing improved performance over existing methods.